摘要
本研究以4岁~5岁儿童认知能力测验为例,在IRT框架下探讨了如何进行追踪数据的测量不变性分析。分析模型采用项目间多维项目反应理论模型（between-item MIRT model）和项目内（within-item MIRT model）多维two-tier model,被试为来自全国的882名48个月的儿童,工具为自编4岁~5岁儿童认知能力测验。经测验水平分析和项目水平分析,结果表明：（1）本文对追踪数据的测量不变性分析方法合理有效;（2）该测验在两个时间点上满足部分测量不变性要求,测验的潜在结构稳定;（3）＂方位题＂的区分度和难度参数都发生变化,另有4题难度参数出现浮动;（4）儿童在4岁~5岁期间认知能力总体呈快速发展趋势,能力增长显著。
Preschool period is an important time for the development of children’s cognition. The most suitable method to analyze the trend of cognitive development is the longitudinal study which shows the development track of children. When researchers analyze longitudinal data, changes in variables are the major focus of research. However, longitudinal data can only be interpreted if important psychometric features of the measured variables are assumed invariant across time. The definition of measurement invariance is that the psychometric features of an instrument remain stable in different situations or at different time points. However, in previous studies, few researchers have considered or tested the measurement invariance of the instrument. This will lower the validity of instrument as the result may be biased or even misleading. The methods to test the measurement invariance include the traditional Confirmatory Factor Analysis（CFA） and the newly developed Item Response Theory（IRT）. Previous researchers always used CFA to analyze measurement invariance. However, comparing with IRT, the CFA method has some limitations in terms of linear assumption of the relationship between observable variables and latent variables. Therefore, the IRT nonlinear assumption is more suitable for reality and psychology issues. Also, the CFA method can only test whether the factor variance and covariance remain stable in different situations, while the IRT method can provide the measurement invariance information at the item level, which is more useful for the future revision of instrument. Thus, we chose the IRT method to systematically analyze the measurement invariance of Children’s Cognitive Development Test. In this study, we selected 882 48-month old children from the country, and the test was conducted in the first year and the second year. After comparing models and considering the instrument construct, we chose the two-tier multi-dimensional item response models for further investigation of the measurement invariance. The testing of measurement invariance included two parts： integral testing and item-level testing. The result indicated that the instrument was partially invariant, but the high correlation between the primary factors indicated that the construct of the instrument remained stable over time. The item-level test results showed that the discrimination parameters and the difficulty parameters of the item ＂direction＂ did not satisfy the invariance condition, which could be considered to be deleted or revised. The possible reason was that children’s conception of direction developed rapidly from age 4 to age 5, and previous studies showed that most 4-year old children hardly distinguish ＂left＂ and ＂right＂. Therefore, the item was difficult for 4-year old children. As the ability of direction grows rapidly, the corresponding task may be too easy for 5-year old children. Moreover, the difficulty parameters of the other four items did not satisfy the measurement invariance condition. This phenomenon was caused by the testing design of this measurement： children accomplished different levels of cognitive tasks at two different time points. The float of difficulty parameters can be tolerated, but the difficulties of these items need to be revised in order to be more suitable for the children of corresponding ages. Additionally, the model estimation results indicated that children’s cognitive proficiency grew rapidly and significantly from age 4 to age 5.
出处
《心理科学》
CSSCI
CSCD
北大核心
2016年第3期606-613,共8页
Journal of Psychological Science
